Linter for LookML
Project description
lookmlint
Lint your LookML.
Provides additional linting checks beyond what is built into the LookML Validator.
usage
From the CLI:
$ lookmlint lint ~/my-lookml-repo
For structured output, set the --json
flag:
$ lookmlint lint ~/my-lookml-repo --json
configuration
lookmlint
looks for a file named .lintconfig.yml
in your lookML project repo.
Its contents can contain lists of abbreviations and/or acronyms you'd like to flag, as well as any checks you'd like to run. More detail below.
sample .lintconfig.yml
abbreviations:
- num
- qty
acronyms:
- aov
- sms
- sku
- sla
timeframes:
- date
- month
- month_name
- time
- year
checks:
- label-issues
- unused-includes
- unused-view-files
- mismatched-view-names
- semicolons-in-derived-table-sql
- missing-view-sql-definitions
- raw-sql-in-joins
installation
Requires python3
.
$ pip install lookmlint
checks
label-issues
acronyms
LookML automatically converts snake case strings (e.g. unit_cost_usd
) to title case (e.g. Unit Cost Usd
), which looks funny when using acronyms. Leverages the list of acronyms defined in .lintconfig.yml
.
Bad
dimension: unit_cost_usd {
...
}
Good
dimension: unit_cost_usd {
label: "Unit Cost (USD)"
...
}
abbreviations
If you'd prefer some words fully spelled out (e.g. 'Quantity' instead of 'Qty'), define a list of abbreviations for lookmlint
to catch.
raw-sql-in-joins
Joins should refer to LookML dimensions as opposed to the underlying fields where possible.
For example:
Bad
join: order_items {
sql_on: orders.id = order_items.order_id ;;
}
Good
join: order_items {
sql_on: ${orders.id} = ${order_items.order_id} ;;
}
missing-timeframes
Find all date/datetime/time dimensions or dimension groups that are missing any of the timeframes defined in .lintconfig.yml
.
unused-includes
If your LookML model explicitly specifies views to include, lookmlint
can catch views that are include
d in your model but not referenced in any of the explorations in that model.
unused-view-files
Find all view files that aren't referenced in any explorations in your project.
views-missing-primary-keys
Find all view files that don't contain a primary_key
dimension.
duplicate-view-labels
Find any cases when two join
s in an exploration end up with the same label.
One way this can unwittingly creep into code is if a label
is defined in a view file, that view is joined twice to the same exploration, but view_label
s are not assigned to those joins.
missing-view-sql-definitions
Find any views that do not have a sql_table_name
or derived_table
value set.
semicolons-in-derived-table-sql
Find any derived table SQL expressions that contain a rogue semicolon, which will throw errors at query time.
mismatched-view-names
Find any views where the view name does not match the view filename.
examples
The sample repo at examples/sample_repo/
contains instances of all linting violations:
~/src/lookmlint $$$ lookmlint lint examples/sample_repo/
Error:
duplicate-view-labels
---------------------
Model: test
Explore: inventory_transfers
Inventory Locations: 2
label-issues
------------
Fields:
View: order_items
- Qty: ['Qty']
- Unit Cost Usd: ['USD']
missing-timeframes
-----------
View: items
Field: Created
- Missing Timeframe(s): ['month_name', 'time']
View: orders
Field: Placed
- Missing Timeframe(s): ['date', 'month', 'month_name', 'time', 'year']
mismatched-view-names
---------------------
- items.view.lkml: order_items
missing-view-sql-definitions
----------------------------
- order_items
raw-sql-in-joins
----------------
Model: test
Explore: orders
order_items: orders.id = order_items.order_id
semicolons-in-derived-table-sql
-------------------------------
- products
unused-includes
---------------
Model: test
- web_sessions
unused-view-files
-----------------
- legacy_products
- web_sessions
views-missing-primary-keys
--------------------------
- order_items
adding to CircleCI
We use CircleCI at Warby Parker to run our checks.
Adding the following contents to .circleci/config.yml
in your LookML project should work for running linting as part of your CI/CD workflow. This all runs in a few seconds, but leveraging caching could also help to speed things up.
Customize the list of checks you run to suit your team's needs, or leave out the --checks
flag to run all possible lint checks.
version: 2
jobs:
build:
docker:
- image: circleci/python:3.6.2-stretch
working_directory: ~/repo
steps:
- checkout
- run:
name: Install lookmlint
command: |
sudo pip install lookmlint
- run:
name: Lint lookml
command: |
lookmlint lint . --checks label-issues,unused-includes,unused-view-files,mismatched-view-names,semicolons-in-derived-table-sql,missing-view-sql-definitions
issues?
This repo is still in alpha, so use at your own risk!
Please open an issue for any feature suggestions or bugs, or feel free to open a PR with a fix / feature!
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for lookmlint-1.0.3-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | f33b37596dfd40b2e981606a79572282c60420439d23c68278db3dab8bb01ab3 |
|
MD5 | 87e8095aa28e345ce72c8a1d22618558 |
|
BLAKE2b-256 | e511f4c32ec77476c42c418182bd343651f03a5460d44ddbb9f049740ddb1dae |